Partnership NHS Trust Records Preliminary Project Report Dr Nicole Baur

The following report is an update on the above project, covering the time period from November 2008 to 20/02/2008. It outlines briefly the current stat of the file deposit and the way of entering the data. The main part of the report shows some preliminary results gained from the data entered so far, and the final part discusses some methodological problems.

1. The deposit As far as the number of patient files in our deposit is concerned, we have to increase the original estimate of 10,000 by approximately 2,000. This is due to the arrival of a second deposit of around 1,000 files last year, all of which have already been entered into the DB. It has also been found that the first deposit contains files from Digby Hospital. These are much thinner than the files, i.e. there are more to one shelf. Thus, the total number of files in our project is likely to be around 12,000. These Digby files have increased our admissions considerably, as, contrary to Exminster, it was custom in Digby to keep all admissions of one patient in one pocket, i.e. the deposit contains files with up to 9 admissions in one pocket. This affects the data entry process, as it takes more time to deal with one file.

2. Data entering The database experienced one final change at the beginning of 2009. The new structure can be seen in appendix 1. It can now store a maximum of 143 pieces of information per patient in 12 related tables. On 5 January 2009, Eliza Newton started as a new project member, employed four days per week, to speed up data entry. The table below shows the number of files dealt with every day. The redesign of the database after Christmas and the creation of data entry forms have facilitated work enormously, and the current progress rate is about 195 files per week. At the current rate the project would need to continue for about another eight months in order to enter all the patient files. Both employees will be available for the next 2 months. Then Eliza will take over sole responsibility for data entry. Her current rate is around 60 files per week, which makes it rather difficult, if not impossible, to complete the project in the proposed time schedule.

Week / c No of files dealt with 12/01/2009 204 19/01/2009 258 26/01/2009 200 02/02/2009 168 09/02/2009 180 16/02/2009 160 Total 1170

Work between the two project members has been organised according to a division of the deposit, starting from both ends. Eliza has worked her way up from 1971 to 1966, dealing mostly with Exminster files. I have entered Exminster files up to October 1951 and am now working on Digby files, which have been put in up to July 1960. While this division keeps interference and overlaps to a minimum, it means that the DB now shows a gap for the mid- 1950s to the late 1960s.

1 3. Database – some results (as of 20/02/2009, if not otherwise indicated) 1

Time period covered in deposit 06/1870 to 10/1971

Patients admitted total 4,517 male 1,758 female 2,759 (= 61.1%) Criminals total 83 male 59 female 24

Most crimes committed fall into the category of petty crimes, i.e. breaking into houses, stealing and shoplifting. Six in-patients were convicted of indecent behaviour and / or assaulting girls, and three of attempted suicide.

Admissions total 6,792 male 2,601 female 4,191 (= 61.7%) From the patient’s status upon admission, which is available for 5,695 admissions, it can be seen that half of all admissions were compulsory (certified, temporary or sectioned). With 63.0% the proportion of compulsory female admissions is slightly above the proportion of female admissions.

1%

25%

37%

2%

11%

24%

voluntary boarder certified informal sectioned temporary voluntary

a) Age on admission average age upon admission: 52.6 years (range: 4 to 101 years) average age upon first admission: 52.1 years average age upon second admission: 53.1 years average age upon third admission: 49.0 years average age upon fourth admission: 50.4 years

These figures suggest that patients with multiple admissions start their “hospital career” in younger years than patients with only one or two admissions.

1 All figures in this report have to be interpreted with some caution due to the current gap in records between 1952 and 1965 explained above. 2 The complete age distribution upon admission is as follows

Age upon Total number of Number (%) of Percentage of admission admissions female admissions female admissions < 10 8 1 12,5 10 – 14 10 3 30,0 15 – 19 149 74 49,7 20 – 24 324 158 48,8 25 – 29 439 234 53,3 30 – 39 933 533 57,1 40 – 49 915 582 63,6 50 – 59 901 590 65,5 60 – 69 1,109 709 63,9 70 – 74 583 324 55,6 75 – 79 426 262 61,5 80 – 84 289 207 71,6 > 84 214 157 73,4

With the data entered so far, the number of admissions is highest for the age group of 60 to 69 year olds. The proportion of female admissions peaks in the older ages, i.e. 80 and over, followed by women in their fifties, sixties and forties. This can be partly explained by the longevity statistics and corresponds with the imbalance towards women in old age generally. Explaining why women in their forties and fifties are more frequently admitted than men is more difficult. It might be to do with the “change of life” many women experience during that age with the children leaving home and working women taking early retirement. It is also an age during which many people have to deal with the loss of close friends and relatives. b) Admissions by year see appendix 2 A significant increase in admissions can be seen from 1949 onwards. Whether this is related to the aftermath of WWII or the foundation of the NHS remains unclear at the moment.

Re-admissions Admissions and re-admissions are affected by the bombing of Digby Hospital in November 1942 and its following closure until November 1947 (patients who were transferred to Exminster or and later re-transferred to Digby had to be entered as new admissions) a) The frequency of admissions is as follows:

Number of admissions Number of patients Number (%) of females 1 3,316 2,000 (60.3%) 2-4 1,071 676 (63.1%) 5-9 111 69 (62.2%) 10-19 12 8 (66.7%) 20 and more 1 1 (100%)

The highest number of admissions of one patient is 35. Interestingly, up to 9 admissions, the ratio male-female admissions mirrors the male-female patient ratio. Ten and more admissions seem to involve women rather than men, but we have to be aware of the small numbers we are dealing with here. b) Place of residence A look at the place of residence (home address) of patients who were admitted repeatedly (see table below) unsurprisingly shows that most patients come from . Nevertheless,

3 neighbouring counties such as Cornwall or sent repeatedly patients to the Devon institutions.

Place of Number of admissions residence 2 3 4 5 6 7 8 9 10+ unknown 31 8 2 1 1 1 Eire 1 Cornwall 8 1 1 1 Devon 643 209 113 52 15 16 8 7 13 Dorset 3 1 1 Hampshire 2 1 Kent 2 1 Lincolnshire 1 8 1 1 Middlesex 1 NFO 5 1 Scotland 1 1 Somerset 9 3 2 Staffordshire 1 Surrey 1 Sussex 2 Wales 2 1 Yorkshire 1 1 1

Appendix 3 shows the place of residence of Devon patients who were readmitted. This table mirrors the one discussed above on a smaller geographical scale. It shows that while patients who were readmitted once or twice came from a wide catchment area, this radius narrows considerably for multiple admissions. These live mainly in areas close to Exeter, particularly along the south coast, as well as in the few major cities in Devon.

Duration of hospitalisation average 2019 days (approx. 5.5 years) Before 06/1948 4,622 days (approx. 12.7 years) From 06/1948 231 days (approx. 7.7 months)

Unfortunately, not all admissions could be included in this calculation, as it required the admission and leaving dates. The calculation was based on 6,208 admissions (91.4% of admissions currently recorded). For a number of (mostly early) admissions, leaving dates were unavailable due to the state of the file. It is therefore likely that the average stay in hospital was longer than our calculation suggests.

Information on how hospitalisation changed a patient’s condition is available for 3,402 of 6,358 discharges. It is interesting to learn that only around 18.5% of patients recovered while in the institution. A much greater percentage improved (31.5%), and 9.4% left the hospital “not improved” – usually against medical advice. On six occasions hospitalisation was unsuccessful, i.e. the patient deteriorated, and 1,373 patients died in hospital. These deaths relate to 30.1% of patients and 20.2% of admissions currently recorded in the DB. According to our data, the success rate of treating patients in the institution does not increase for patients who stayed there longer. It might be interesting to note that 16 patients absconded during their stay in hospital.

4 Geographical information Admissions to all three hospitals were primarily from Devon (Exminster 94.4%, Digby 95.4% and Wonford House 74.4%). Exminster had the largest catchment area with admissions from 25 English counties, Eire, Scotland and Wales as well as four admissions of patients with a foreign address. Digby admitted patients from 10 English counties, as well as Eire and Scotland, while Wonford House patients came from 16 English counties and Eire / Scotland and Wales. The neighbouring counties Cornwall and Somerset were home for patients particularly admitted to Exminster and Wonford House.

Map 1 (will be available at the meeting) shows the place of residence within Devon for patients admitted to the three hospitals. A clear concentration of patients along the southern coastal regions, particularly and is evident. Other areas with large numbers of patients are the bigger cities such as , , , etc. (data as of 30/01/2009)

A map showing the number of admissions related to 10,000 inhabitants of the patient’s area of origin for the year 1951 is currently being produced.

Linkage information With the foundation of the NHS and the subsequent emphasis on community care treatment became more spatially dispersed. This placed a new importance on communication, and the more recent files are full of professional correspondence, some of which might be useful in terms of linking institutional and community care. a) GP Details of the patient’s GP are available for 1,795 patients. Of another 50 patients we know that they were not registered with a GP. Information on the patient’s private doctor is available either from the “Statement of Particulars” or via admission / discharge letters. In case of an admission letter we might have the benefit of learning about previous treatment and medication administered to the patient. Correspondence can indicate that, for example, the GP referred the patient to social services or to the Out-patient Department of the nearest hospital where they can be seen by a consultant who might then recommend their admission. The discharge letter issued by the hospital usually discusses diagnosis and treatment as well as recommending further medication and, if required, out-patient care. b) NI number 338 patients c) After -care reports 340 patients Upon leaving the hospital, patients were usually asked to sign a form stating whether they agree / disagree to after-care treatment. In case of agreement a social worker will be visiting the patient occasionally at home and provide a report. A copy of it can usually be found in the patient files. These reports, which are very variable in numbers, can take any form from a few lines up to a 2-page minute account of the patient’s experience since either discharge from hospital or the previous visit. Occasionally they suggest a patient’s readmission. d) Previous admissions to other hospitals 681 patients We have got 1,015 accounts of hospitalisation in mental institutions before treatment was sought in one of the Devon hospitals. Some patients had been treated up to 13 times in

5 other institutions, and occasionally full accounts of their condition and treatments are available. e) Out-patient treatment 56 patients From about the 1950s onwards, most patients are recommended for out-patient treatment after leaving the hospital. Occasionally evidence of their treatment has been placed with the files. Apart from out-patient treatment there was the option of patients being seen by a consultant psychiatrist before there admission to hospital. These consultations usually took place at a hospital, there is, however, evidence of domiciliary visits by the consultant

It is often difficult to reconstruct what agencies were involved in a patient’s admission / discharge because the files contain mostly carbon copies which do not always show the author’s name, signature or institution. In addition, there seemed to have been a variety of job titles, such as “Duly Authorised Officer / (Senior) Mental Welfare Officer / (Senior) Social Worker in Mental Health / Psychiatric Social Worker”. It remains unclear whether they all had a similar function or in what way they differed. More material on the admission process will be available during our meeting, so suffice it to say that the social / welfare department was usually contacted by either relatives or the patient’s GP, occasionally by the consultant psychiatrist or the police, and that these agencies play a key role in the admission process. f) Other potential linkage material Welfare committee minutes for the period 1948 to 1961 have been examined, but have to be regarded as of limited linkage value because they do not mention individuals. However, we positively identified records of the Exeter Clerk’s records with patients in our DB. These records are more detailed and could fill some gaps in our DB, but they refer only to patients diagnosed as M.D.

The Cornwall Records Offices has been contacted regarding a potential comparison with St Lawrence’s Hospital, Bodmin. They are, in principle, willing to cooperate and make the patient files available for research, provided that we can obtain clearance from the Cornwall NHS Partnership Trust.

Diagnosis a) Suicide 278 patients were regarded as suicidal upon admission with a further 77 being “potentially suicidal”. Of these 355 patients 206 had actively attempted to commit suicide, some on several occasions. 74% of attempted suicides were committed by female patients. 38 patients had threatened to “do away with themselves” prior to their admission. According to our data, the preferred methods of attempting suicide were overdose of drugs and (coal) gas poisoning. Other methods included cutting throats or wrists, drowning and jumping from a window. Only male patients attempted to shoot or hang themselves or walked in front of a car / bus / train. b) Frequent diagnoses The table below shows the proportion of the most frequent diagnosis in our DB (they account for about 89% of diagnoses in male and female patients). Melancholia or later depression was diagnosed in a total of 805 patients, followed by dementia (628), schizophrenia (460), Insanity (323), Psychoses (288) and Insanity (270). Our data suggests that there were considerable differences in male and female patients diagnosed with certain diseases. A much higher proportion of women, for example, was diagnosed with paranoia, mania and

6 melancholia. The higher proportion of female dementia patients is most likely due to the higher percentage of women admitted in old age. Men, in contrast, were more often treated for alcoholism, GPI, insanity, and schizophrenia. In addition, a larger proportion of male patients were diagnosed as psychopathic personalities and mental defectives.

Diagnosis Percentage of male patients Percentage of female patients Alcoholism 2,9 0,2 Anxiety 3,9 3,2 Confusion 5,1 3,4 Dementia 13,0 16,7 GPI 1,5 0,2 Hypomania 1,2 1,0 Hysteria 0,7 1,1 Inadequate personality 0,2 0,1 Insanity 10,4 6,2 Mania & manic-depressive 4,1 8,2 Melancholia & depression 17,2 21,1 Moron / MD / Imbecile 6,9 5,6 Paranoia 0,8 4,6 Psychosis 7,1 7,0 Psychopathic personality 1,3 0,5 Schizophrenia / dementia praecox 12,5 10,3 c) Diagnosis and age The variety of diagnoses recorded in the DB makes it difficult to group them according to the patient’s age, particularly for “middle-aged” patients. It is worth noting, however that all patients under the age of ten were classed as mental defectives. In the category 10 to 14 years, MD was still the most frequent diagnosis, but it was joined by psychopathic personality, epilepsy and schizophrenia. Depression / melancholia and mania were diagnosed in teenagers as young as 15 to 19 years. Together with anxiety, these mental illnesses account for most diagnoses in patients aged 20 to 59. Even patients in their sixties were frequently diagnosed with depression, but from that age onwards dementia takes over as the main diagnosis.

Treatment Treatments administered at the Devon hospitals are in line with existing literature on this topic. Our results are again somewhat affected by fragmentary or missing records, but below are results based on 2,533 patients for whom we currently hold information on various treatments. These fall into three main categories: a) Physical treatment ECT 730 patients More than two thirds (483) of patients treated with ECT were women. This is not surprising given the higher percentage of women diagnosed with depressive illnesses, as ECT was primarily used to treat depression / melancholia (44.2%). It came also to use in patients with schizophrenia (11.4%) and various forms of anxiety (8.1%). To a lesser extend it contributed to the treatment of mania / manic-depressive illnesses as well as psychoses and for mentally defective / morons / imbeciles. Most patients received only one course of ECT (usually six to eight treatments) during their hospitalisation.

7 Insulin Coma Therapy 49 patients Insulin Coma Therapy was mainly administered between 1947 and 1955 to treat schizophrenia as well as some cases of puerperal psychoses and mania. In more recent years, however, a “modified insulin therapy” was used where patients were injected with insulin “until sweating occurs”. This treatment has a wider application in alcoholism, anxiety, dementia, depression, hysteria as well as schizophrenia.

Malaria Treatment 22 patients treated for GPI, delusional insanity and schizophrenia

UV Treatment 5 patients between 1928 and 1936, all suffering from mania (1) or schizophrenia (4).

Leucotomy 5 patients starting in 1959. Leucotomies attracted great interest among the professional community, and documentary evidence suggests that the patients were anxious to undergo this procedure. Patients of the Devon hospitals were referred to Moorhaven () or Frenchay () to have the operation. b) Tonics and mixtures were usually given before the onset of drug treatment, some however survived into the 1960s. c) Psychoactive drugs Sedative drugs such as paraldehyde, sodium amytal chloral hydrate and phenobarb have a long tradition in the Devon hospitals. The discovery of chlorpromazine in 1952, however, started the widespread use of psychoactive drugs. It is interesting to see that even patients who had been in hospital previously without any drug treatment, were put on one of the new drugs upon readmission. The two main categories of drug used for the patients in our deposit were anti-psychotics and anti-depressants. In the beginning there seemed to have been a lot of confusion as to the right dosage. Patients had their dosages changed frequently, and some were put on extremely high dosages, such as for example 400 mg daily (Largactil). In more recent years dosages decreased. Below is a list of the most frequently used drugs and the date when they were first used for our patients:

Anti-psychotics: Largactil May 1953 Fentazin October 1958 Melleril February 1959 Serenace August 1959 Taractan September 1960 (discontinued in 1962)

Anti-depressants: Tofranil January 1959 Nardil August 1959 Marplan February 1961 (discontinued 1962) Tryptizol May 1961 Pertofran February 1963 Surmontil April 1966 Prondol October 1968

8 Despite these new drugs, barbiturates continued to play a major role in the treatment of mentally ill patients. Almost every patient was put on a sedative, if only for a few days after admission.

It might be interesting to know that some of the drugs administered to our patients, such as Nembutal or Drinamyl, have since been taken off the UK market due to their severe adverse effects.

4. Methodological problems As has already been hinted at in this report, there are some methodological problems related the data itself as well as to our options of manipulating and interpreting it.

Structure and content of the deposit a) We currently have potential duplicates in the tables “patient” and “gp”, as we have too little information to verify data (e.g. some doctors have gone into the table with two addresses, as it was impossible to find out whether they have moved or if they are the same people at all). Census data or a register of addresses / professions might help here b) Only a small number of files is available for the years 1945 – 1949. It might be good idea to check admission books for annual admissions & then compare how many of them we have got in our DB c) Wonford House: hardly any files available, unless patient was also treated in Exminster & Digby. The DB contains records of Wonford patients up to 1944, which have been entered from Hospital Index Cards. Is it worth entering information from the remaining HICs, even if we are unlikely to encounter any files (DRO intends to hold on to all HICs)

Demographic data a) The deposit contains a very large number of elderly people from about 1960 onwards. As one of the researchers has started from the end of the deposit, this can cause a slight imbalance in the average age until all files have been entered b) If we want to analyse patients’ occupations, these need to be classified. A model, based on the HISCO classifications is provided on the following website on the history of work: http://historyofwork.iisg.nl/major.php . Its suitability for our data will have to be evaluated in more detail.

Geographical analyses a) A clear agreement on names of counties (before / after the reform) is vital to any geographical analysis. Particular problems are caused by e.g. Lyme Regis which must have switched between Devon & Dorset, and by whether we want to consider Plymouth a part of Devon b) I was able to obtain from the DRO a map of Devon and to import it into a graphics programme as well as a population count for these (census data 1801-2001). These tools allow a graphic representation of e.g. the patients’ places of residence. If we want to use a GIS instead, we have to be aware that the patient files give full addresses minus post-codes (I might be wrong, but I assume the GIS maps are post-code based). As we are dealing with a very long time period, it would also be useful to know if the university (probably Geography) has digitised large-scale maps for earlier periods and whether they would allow us access to them

9 Drug treatment The collection of data on drug treatment remains the most difficult task of the project and might require some assistance from someone with professional experience in this sector. Specific problems are: a) Names of drugs on Medicine Cards are often illegible / abbreviated (e.g. Hst Par = Paraldehyde?); many of the older drug names cannot be found on the internet b) The duration of the treatment is not always available; we usually know when patients commenced a particular drug treatment, but whether the dosage changed or when it finished is not always evident from the files c) Apothecaries’ measures and weights are difficult to decipher, particularly the difference

between “ounce” and “drachme”, as the handwriting is often open to wide interpretation...

10 Appendix 1 : Structure of the database (as of 05/01/2009)

11 Appendix 2 : Admissions by year

Year Number of admissions 1929 103 1870 2 1930 95 1875 1 1931 69 1881 1 1932 94 1883 2 1933 83 1886 1 1934 108 1887 1 1935 95 1890 2 1936 123 1891 2 1937 129 1893 1 1938 132 1894 1 1939 135 1895 2 1940 116 1896 1 1941 130 1897 3 1942 156 1898 1 1943 135 1899 7 1944 162 1900 9 1945 61 1901 4 1946 78 1902 8 1947 75 1903 4 1948 95 1904 8 1949 256 1905 9 1950 427 1906 6 1951 349* 1907 13 1952 147 1908 7 1953 102 1909 9 1954 49 1910 20 1955 64 1911 19 1956 60 1912 15 1957 82 1913 17 1958 179 1914 12 1959 385 1915 11 1960 155 1916 19 1961 163 1917 20 1962 169 1918 32 1963 33 1919 32 1964 17 1920 17 1965 25 1921 28 1966 25 1922 36 1967 147 1923 36 1968 278 1924 45 1969 432 1925 58 1970 286 1926 55 1971 95 1927 47 1928 69

* Years 1951 to 1971 have not been fully entered yet

12 Appendix 3 : Place of residence of readmissions from Devon

Place of residence Number of admissions 2 3 4 5 6 7 8 9 10 11 12 14 15 17 19 35 1 1 Appledore 2 1 Ashburton 5 1 Ashford 1 1 6 Bampton 2 Barnstaple 17 7 4 1 1 2 1 Beer 1 1 Bestone 1 1 2 1 Bickleigh 1 Bideford 13 4 4 1

Bishopsnympton 3 Bishopstawton 1 1 1 1 1 Bovey 2 1 1 Bow 2 1

Brampford Speke 1 Bradford 1 1 1 1 1 1 1 Bratton 1 1 9 1 1 1 16 2 2 1 1 1 1 4 1 1 1

Buckland Brewer 1

Budleigh Salterton 3 1 Burrington 1 3 1 1 1 Cheriton Bishop 1 1 1 1 Clovelly 1 1 1 1 1

13 1 Churston 1 1 1 Colyton 3 1 3 1 Crediton 9 3 1 1

Croyde 1 4 3 1 1 1 1 Dartmouth 7 1 1 13 3 2 2 1 Devonport 1 1 Dolton 1 1 1 East Ogwell 1

East Portlemouth 1 1 Exeter 137 38 26 13 4 7 1 1 Exminster 2 19 13 6 2 2 1 Exton 1 Faringdon 1 1 2 Fremington 1 1 1 1 Halwell 1 1 1 1 Hartland 1 1 1 2 1 1 High Bray 1 Holsoy 1 4 7 3 Hooe 1 1 Ide 1 9 1 1 3 1 1 1 1 Ivybridge 1 Kenn 1 Kennford 1 1

14 1 Kenton 2 1 1 1 2 1 1 1 1 Kingsnympton 1 5 1 Kingston 1 1 3 1 1 Lewdown 1 Lifton 1 1 Lower 1 2 2 1 1 2 1 1 2 1 Membury 1 Merton 1 1 2 1 Monkton 1

Morchard Bishop 2

Moretonhampstead 1 1 3 1 1 23 7 9

Newton Poppleford 1 1 1 Newton Tracey 1 1 1 3 Northam 2 1 2 3 2 1 1 3 2 3 2 Paignton 28 13 4 3 2 1 1 1 2 1 Plymouth 6 3 2 Plympton 8

Plympton St Maurice 1 1 1 1 1 1 15 1 1 1 1 1 Sandford 1 1 Seaton 4 4 1 1 1 Shute 1 22 4 2 2 1 1 Silverton 4 1 1 4 1 1 7 1 1 1 1 1 5 2 1 1 1 1 Tamerton Foliot 1 3 2 3 1

Tedburn St Mary 1 1 13 5 1 1 1 1 Tiverton 11 4 2 2 1 Topsham 10 4 2 1 Torquay 49 24 9 2 2 2 1 Torrington 3 2 7 2 1 1 1 2 1 1 2 1 1 1 1 1 1 1 Westleigh 1 1 1 1 1 Westward Ho! 2 1 Witheridge 1

Withycombe Raleigh 1 Woodbury 1 1 1 Yelverton 2 1

16